Optimization of recharge and pumping rates by means of an inverse 3D model

Abstract

The paper concerns the optimization of pumping and artificial recharge rates under hydraulic head constraints. This optimization is accomplished by means of an inverse three-dimensional (3D) model which is a combination of a numerical model, sensitivity analyses and nonlinear regression. The optimization criterion used is the minimization of the sum of the squares between the calculated and the desired water table maps. These maps were established for the end of each month of the period from December to August so that the annual average flux is outward from the modeling area. This outward flux will push back the salt water intrusion. The assumed natural recharge rates of the different months equal the mean values of the period 1958–1977. The aquifer parameters describing the water conducting and storing properties are also required as input data as well as the initial hydraulic heads. The optimized parameters are the artificial recharge rates of the months December, January, February and March and the pumping rates of the months April, May, June, July and August. For the first-mentioned months, the pumping rates are considered as known. During the last-mentioned period, no artificial recharge is possible. A unique solution is obtained. To check the optimized rates, a water balance is made. With this balance, it is shown that the treated phreatic aquifer can be used as a subterranean reservoir to store water that is abundantly present in the winter until the summer when water demand is large. It is also proved that the salt-water intrusion can be stopped by the temporal storage of the artificial recharged water.